In the last article I talked about AI and searching for stuff in big companies. Employees will be able to use AI to get answers drawn from all the different places that information is stored – files, databases, email and calendars, Teams chats, all the rest. The AI will only get information that you have permission to access.
I mentioned that this will be good for Microsoft.
The reason is that access to all those different servers is almost certainly run by Microsoft. When you log in to the company network, it’s Microsoft programs that confirm your identity and decide what you can and can’t access.
The real money in today’s world is made by selling services to rich companies. Remember the old story about getting rich in a gold rush by selling picks and shovels to the miners? Microsoft provides fundamental IT services – 21st century picks and shovels – to virtually every company on earth. And that gives Microsoft a crucial head start for integrating AI into the businesses.
TRIGGER WARNING – MILDLY GEEKY WORDS. The core component, the plumbing that runs large businesses, is Azure Active Directory. Actually now it’s called Entra ID because Microsoft hasn’t met a brand name that it doesn’t change once a month. In any case, that’s just one underlying part of a vast Microsoft ecosystem to verify user identities and control what they can access.
More than 90% of Fortune 500 companies use Microsoft services for at least some portion of identity and access management on company networks. Microsoft has similar penetration in businesses of all sizes. There are competitors and complicated relationships but it’s a fair simplification to say that Microsoft programs are in control of access to most business networks.
Now Microsoft wants to leverage its position and be the company that handles AI for businesses.
Microsoft is working frantically to embed AI into its products so companies can deliver personalized answers to questions in business that draw from all the data everywhere in the company at the same time, but do it in a way that respects user permissions.
When Microsoft does that – and it will – employees will have tools for gathering information that are light years ahead of what can be done today. And the company will continue to have data security and compliance.
Let’s quickly do one more example. Imagine you work at a healthcare company that uses a program like Epic or Veeva to store HIPAA protected patient data.
Microsoft software almost certainly handles your login to the company network.
Once you’re signed in, Microsoft tells Epic or Veeva who you are. Those systems take over control of what you’re allowed to access in their databases based on your role as doctor, nurse, researcher, admin staff, etc. The goal is to avoid multiple login names and passwords. It’s called “Single Sign-On” and it’s all the rage in enterprise IT circles.
When AI answers questions, it will respect those permissions. Perhaps it will be Microsoft running the AI for searches in company documents, maybe questions in the Epic/Veeva systems will be handled by the AI built in by those companies.
But Microsoft will be the gatekeeper and the money flowing to Microsoft from enterprises will grow and grow and grow.
Don’t underestimate Microsoft’s laser focus on enterprises. Remember, Microsoft no longer has a division named “Windows” because Windows isn’t important enough to justify it any more compared to the enterprise products. (Seriously. No joke.) Now Microsoft wants to run enterprise AI and leverage its success in controlling identity management.
When you log into a company, the AI will know who you are. And Microsoft will get another few dollars in licensing fees. Ka ching!